M3 - Hypotheses Flashcards
Goal of ER in Mgmt & Economics
- developing theories & hypotheses
- testing hypotheses
- measuring strength of effects
- predicting economic developments
Hypotheses are…
Assumptions about structural properties of reality.
They point behind an individual situation and can be disproved by empirical data.
They have to …
Have general validity
Be falsifiable
Be frased in conditional clauses or equality/inequality
Criterias for hypotheses:
- general validity
- falsifiable
- frased as conditional clause or show equality/inequality
Falsificatot
A falsificator is a possible data combi that leads to the rejection of the hypothesis.
The more falsificators, the higher the info content! 👍
!use H0 to falsify, because H1 can never be falsified!
Why use falsification rather than verification?
Before a prediction has been statistically verified, it must be considered as incorrect.
Only hypotheses that have repeatedly not been rejected will be maintained as ‘proven hypotheses’/ confirmed
Statistical hypotheses
- are only valid in ……..
- refer to the …. we examine
- are defined as assumptions about …..
- make …… ……
- use a …. selection out of the ….
Statistical hypotheses
- are only valid in the majority of cases
- refer to the population we examine
- are defined as assumptions about the distribution
- make probabolistic statements
- use a random selection out of the population
3 types of hypotheses
- association (a & b are related)
- causality (a acts causally upon b)
- difference (a is different from b)
–> causality does not mean correlation!
Alpha error
H0 is true, but is rejected
B-error: H0 is wrong, but is accepted
Beta-error
H0 is wrong, but is accepted.
A-error: H0 is true, but is rejected
The probability of error
Probability of making an error
Hypothesis
full model
restricted model
full model has all in it
restricted model: force the intercept to be zero
–> decsribes the data less (regr line goes through zero)
